Filling a softbox with a speedlight

Messages
23,200
Name
Richard
Edit My Images
No
Following on from this other thread https://www.talkphotography.co.uk/threads/soft-octa-box-for-speedlite.655928/ here are some comparisons showing different options when using a speedlight flashgun in a softbox. The objective is to fill the softbox all over as evenly as possible, which is more difficult with speedlights that project all the light straight forward, as opposed to studio strobes with a bare-bulb flash tube that spreads the light in all directions.

To cut to the chase, it's a popular claim that fitting a Stofen-type diffuser cap mimics a bare-bulb for the best result. The comparisons here show that this is not the case, and that actually the inner diffuser panel already supplied with most softboxes is also effective with speedlights (set to the widest zoom-head setting, without the wide-panel).

Furthermore, there's a big drop in brightness with a Stofen - 1.6 stops here. Using the wide-panel is similar - a drop of 1.7 stops. In comparison, fitting the inner diffuser panel reduces brightness by only 0.3 stops.

The softbox used was a 90x90cm square Pixapro. This one http://www.essentialphoto.co.uk/product/pixapro-90x90cm-square-5cm-grid-softbox/
Flash units were an Elinchrom D-Lite studio strobe, and Yongnuo 600EX-RT speedlight.
Godox speedlight bracket https://www.amazon.co.uk/gp/product...mini_detail?ie=UTF8&psc=1&smid=A2EFUPWDJIRSGN
The Stofen-style diffuser cap used was this one http://www.ebay.co.uk/itm/Flash-Bou...61480649?hash=item419a4affc9&autorefresh=true and the 'shower cap' is here https://www.premier-ink.co.uk/photographic/universal-soft-flash-diffuser.html

There are seven sets of comparisons, presented as seven separate posts due to the site files limit - click for full size. Each includes an actual image taken with the softbox (at 1.0m distance, balanced for exposure and colour in PP) and a shot of the front of the softbox. On many bloggers' websites, images like these can be very misleading unless the brightness is properly 'pegged' to the same values, so there is also an analysis graphic that shows true light levels (Edit: see post #10).

The seven options are:
1) D-Lite with both diffuser panels (best result, for reference)
2) D-Lite with front diffuser only
3) YN600 at 20mm with front diffuser
4) YN600 with 14mm wide-panel and front diffuser
5) YN600 at 20mm with Stofen and front diffuser
6) YN600 at 20mm with shower-cap and front diffuser
7) YN600 at 20mm with both diffuser panels

There are some slight differences visible in the shadows, but they're hardly dramatic! See what you think :)

1) D-Lite with both diffuser panels (best result, for reference)
D Lite double diff-3379.jpg D Lite double diff-3421.jpg D Lite double diff-3421_f-stop_ctrY.png


 
Last edited:
2) D-Lite with front diffuser only
D Lite single diff-3380.jpg D Lite single diff-3416.jpg D Lite single diff-3416_f-stop_ctrY.png
 
3) YN600 at 20mm with front diffuser
YN 20mm single diff-3382.jpg YN 20mm single diff-3413.jpg YN 20mm single diff-3413_f-stop_ctrY.png
 
Last edited:
4) YN600 with 14mm wide-panel and front diffuser
YN + wide panel single diff-3386.jpg YN + wide panel single diff-3409.jpg
 

Attachments

  • YN + wide panel single diff-3409_f-stop_ctrY.png
    YN + wide panel single diff-3409_f-stop_ctrY.png
    76.9 KB · Views: 18
Last edited:
5) YN600 at 20mm with Stofen and front diffuser
YN + Stofen single diff-3388.jpg YN + Stofen single diff-3407.jpg YN + Stofen single diff-3407_f-stop_ctrY.png
 
Last edited:
6) YN600 at 20mm with shower-cap and front diffuser
YN + shower cap single diff-3390.jpg YN + shower cap single diff-3403.jpg
 

Attachments

  • YN + shower cap single diff-3403_f-stop_ctrY.png
    YN + shower cap single diff-3403_f-stop_ctrY.png
    75.9 KB · Views: 15
Last edited:
7) YN600 at 20mm with both diffuser panels
YN + double diff-3394.jpg YN + double diff-3399.jpg YN + double diff-3399_f-stop_ctrY.png
 
Just got time for testing output and having only a 60cm octa a comparison wont make sense.
On my TT685 in octa initial trial shows applying both the wa reflector and stofen type diffusor costs less than 1 stop. Ill do a propper test tonight.

Edit.
Now before you start scratching your head let me add what I forgot in the rush between cerials and coffee, packing lunchboxes, setting up and writing the above. The octa is an umbrella type with the mechanics blocking straight ahead from the flash, thats why. So different equipment different findings. :)
 
Last edited by a moderator:
  • Like
Reactions: TG.
Just got time for testing output and having only a 60cm octa a comparison wont make sense.
On my TT685 in octa initial trial shows applying both the wa reflector and stofen type diffusor costs less than 1 stop. Ill do a propper test tonight.

Edit.
Now before you start scratching your head let me add what I forgot in the rush between cerials and coffee, packing lunchboxes, setting up and writing the above. The octa is an umbrella type with the mechanics blocking straight ahead from the flash, thats why. So different equipment different findings. :)

Apologies for an error - loss of brightness with the Stofen is 1.6 stops*. My silly mistake when adding/subtracting ISO and Lightroom tweaks and going the wrong way. That seems more like it! I've corrected the first post.

That brings it more in line with your findings, and I would echo your comment that different equipment gives different results. From these tests, I think the takeaway findings are that a double-diffuser is a good and efficient method with speedlights in softboxes, as it is with bare-bulb strobes - which is what I've found on several occasions with different set-ups. Stofens are no better and lose brightness, ditto the wide-panel. Note that I didn't test a Stofen with the double-diffuser. But trying to analyse everything down to the last tenth of a stop is taking things too far.

What I think is happening here is the inner diffuser is also a reflector - half the light passes through, half is bounced to the back of the softbox that is now acting like a reverse-firing softbox - and they are known to work well with speedlights. The rear of the softbox is important and I have cut a silver cardboard surround that covers the big black gaps that you normally get around the (Godox) speedlight bracket.

*Edit: exposure differences are taken from the pictorial images of the manikin
 
Last edited:
Thanks for posting this. This was just what I needed to know. I only have speedlights at the moment but this helped me loads.
 
Even the best case example doesn't look that even, would it make more sense to use a better designed softbox? That way you can see exactly how good or bad each head is.
 
The square grey images of the softbox front have been adjusted for brightness visually, using the mid-tones towards the corners as reference. This seemed like the best way, but whatever reference is taken, there is a lot of scope for misleading results. You can basically adjust them to show anything you want, and if the highlights are used as a reference for example, some of them would be very dark away from the centre.

The best result is not the one with the biggest bright patch, but the ones with the most even grey tone overall - and that's delivered by the double-diffuser both with speedlights and bare-bulb strobes. The double-diffuser is also more efficient in terms of exposure than a Stofen, which is extra important given the low power of speedlights.

The other notable result is how very similar the pictorial images are of the manikin. Basically the unevenness on the front of the softbox doesn't translate to uneven lighting across the subject, because all areas of the subject receive light from all parts of the sofbox, and the uneveness is homogenised. That's not to say it doesn't matter, but in terms of the final images here it makes much less difference and is much less important than you might imagine :)
 
Last edited:
Even the best case example doesn't look that even, would it make more sense to use a better designed softbox? That way you can see exactly how good or bad each head is.

This was done to dispel a rumour, other interpretations are also available. Most soft boxes will give a similar result understanding what @HoppyUK has done here, more to the point it needs a better light

Mike
 
Even the best case example doesn't look that even, would it make more sense to use a better designed softbox? That way you can see exactly how good or bad each head is.

It's a standard size 90cm square double-diffuser softbox, 90 degrees angle between the sides, average depth. Pretty representative I'd say. There are thousands of possible combinations and permutations - feel free to demonstrate a few more ;)

The part of the softbox that I think is most overlooked and has a significant influence on shaping the light that comes out of the front, is right at the back. Along these lines, I've spent some time recently comparing speedring mounts, particularly with the Godox AD600 head that has a more recessed flash tube than other Bowens-mount strobes (@ Mike Weeks ;)) Use this chromed DynaSun speedring adapter http://www.ebay.co.uk/itm/262991993062?_trksid=p2060353.m2749.l2649&ssPageName=STRK:MEBIDX:IT and mount it in front of the speedring rather than to the back, and you'll gain half a stop of brightness compared to this much more common dull grey alloy variety that's much deeper http://www.ebay.co.uk/itm/Inner-Bow...241288?hash=item25ca9f82c8:g:VHcAAOSwEzxYeKzu It looks like a very small difference, but half a stop is worth about 150Ws with that head.

Edit: maybe should have used an octa rather than a square softbox, to reduce that square cross pattern, or rotated the softbox to a diamond (point at the top) so the flat sides didn't align with the square speedlight flash tube. Would have changed the light pattern on the front of the softbox, but not the pictorial result.
 
Last edited:
As a sidenote
I read somewhere the ad600 was not originally designed for the bowens mount that came along as an afterthought and it does compromise the flash a bit due to placement of the bulb in the modifier.
 
Last edited by a moderator:
As a sidenote
I read somewhere the ad600 was not originally designed for the bowens mount that came along as an afterthought and it does compromise the flash a bit due to placement of the bulb in the modifier.

Yes, true.

Off topic, but and they've just stuck a Bowens mount on the front without repositioning the flash tube that needs to be 10-15mm further forward - as Bowens-brand heads are. That would obviously increase the overall length, but without that the flash is firing out of a little tunnel, the light can't distribute properly and gets buried in the mount. I have a few other Bowens-fit heads here ATM and Godox is not the only one to benefit from a low-profile softbox adapter, but the AD600 is notably the worst affected.

Edit: a little dish reflector inside the mount also makes a big difference, helping to project all the light forward. I've made one (out of a carboard party bowl and chrome tape) that gains about a third of a stop, though the better answer is a low-profile (and shiny!) adapter.
 
Last edited:
I'm rather surprised by the pattern of light coming from the D-lite; getting that pattern from a speedlight is completely expected, but not from a ring tube flash. That suggests to me that for some reason the softbox is focusing the light, and more-so in one azimuth. While I don't think that significantly affects *these* results/conclusions, it might lead one to draw the conclusion that a forward firing speedlight is just as even as a bare bulb flash. I don't think I would say that is generally the case, although it apparently "can be."
 
Last edited:
I'm rather surprised by the pattern of light coming from the D-lite; getting that pattern from a speedlight is completely expected, but not from a ring tube flash. That suggests to me that for some reason the softbox is focusing the light, and more-so in one azimuth. While I don't think that significantly affects *these* results/conclusions, it might lead one to draw the conclusion that a forward firing speedlight is just as even as a bare bulb flash. I don't think I would say that is generally the case, although it apparently "can be."

I think what we're seeing is the result of a square softbox with a square inner diffuser, and a speedlight projecting an oval pattern that's obviously stronger left/right than it is top/bottom.

With the Elinchrom (and most other heads too) there are brighter spots towards the left and right sides. That particular softbox has two little flaps either side of the mount - see pics here http://www.essentialphoto.co.uk/product/pixapro-90x90cm-square-5cm-grid-softbox/ They're for ventilation or inspection or something, but I don't use them so I've covered them over with chrome Duck tape inside, but they do very slightly change the profile of the softbox sides at that point and are flatter and less curved. It looks like they're at just the right angle to pick up a reflection of the flash tube. The Duck tape has a slightly stippled surface, but maybe that needs to be more irregular to break up the reflection more, or I might try to put a bit of a curve in them.
 
A very interesting and educative thread, many thanks for it.
Even the best case example doesn't look that even, would it make more sense to use a better designed softbox? That way you can see exactly how good or bad each head is.
Up to a point I agree with you, but there are different interpretations of this
1. Many (most?) hotshoe flashgun users do go for really cheap softboxes, the people who take softbox design and quality seriously usually either go for portable flash systems or studio flash. And better, deeper softboxes are often considered to be "too bulky" by hotshoe flash users.
2. Poorly designed softboxes, made with poor materials that aren't taut even when new bounce light around in all sorts of different directions, and this can actually help.

Also, for many of the people who use hotshoe flashguns in softboxes, it really doesn't matter, if people are photographing shiny products with complex shapes, it makes an enormous difference whether the lighting is perfectly even or not, but these photographers tend to be specialists who have the equipment designed for their area of expertise, most people who use hotshoe flashguns with softboxes use them for far less demanding work. And many people use them just to add a bit of fill light (which is the best that they can do anyway in fairly bright lighting conditions) and nobody using a flashgun just to add a bit of fill will see any any difference, because any differences in the eveness of the light will be theoretical and won't be visible.

The point about the recessed flash tube on the Godox AD600 is a good one. If you want to buy a flash that does its job well, get it from a manufacturer who understands about lighting, not from Godox:) Does anyone remember that awful flash head that Yongnuo introduced? The flash tube on that one was so deeply recessed that it was unusable...
 
I think what we're seeing is the result of a square softbox with a square inner diffuser, and a speedlight projecting an oval pattern that's obviously stronger left/right than it is top/bottom.
I don't think I've ever seen a strobe that was completely even, there's always some kind of pattern to it. It just surprises me that a ring tube would project such a strong horizontal "bar" pattern, almost like the diffusion was curved/folded towards the head (but it doesn't look like it was). I would hope that's not actually characteristic of the D-lite...
 
It's a standard size 90cm square double-diffuser softbox, 90 degrees angle between the sides, average depth. Pretty representative I'd say. There are thousands of possible combinations and permutations - feel free to demonstrate a few more ;)

I'm just a bit surprised by the pattern I saw like Steven mentions, shouldn't a square softbox produce a cross pattern rather than a single band with a standard exposed tube head? Gary makes a good point that people are going to be using that kind of softbox but it just caught me by surprise as it wasn't what I'd expect. Here's my 90x90 for reference:

DSC_0087_1.jpg

I've been spoiled lately by rear firing octas, I'm guessing they'd eat too much light to be useful for a speedlight though (and far too big for convenience).
 
Last edited:
The internal of my softbox
IMG_20170609_184413~2.jpg
Note no inner diffusor
Flash set at 24mmDSC00010.JPG

Wide angle reflector down (14mm)DSC00011.JPG

Wide angle diffusor and "stofen" diffusor
DSC00012.JPG

@24mm and stofen attached
DSC00013.JPG
 
Last edited by a moderator:
A very interesting and educative thread, many thanks for it.

Up to a point I agree with you, but there are different interpretations of this
1. Many (most?) hotshoe flashgun users do go for really cheap softboxes, the people who take softbox design and quality seriously usually either go for portable flash systems or studio flash. And better, deeper softboxes are often considered to be "too bulky" by hotshoe flash users.
2. Poorly designed softboxes, made with poor materials that aren't taut even when new bounce light around in all sorts of different directions, and this can actually help.

Also, for many of the people who use hotshoe flashguns in softboxes, it really doesn't matter, if people are photographing shiny products with complex shapes, it makes an enormous difference whether the lighting is perfectly even or not, but these photographers tend to be specialists who have the equipment designed for their area of expertise, most people who use hotshoe flashguns with softboxes use them for far less demanding work. And many people use them just to add a bit of fill light (which is the best that they can do anyway in fairly bright lighting conditions) and nobody using a flashgun just to add a bit of fill will see any any difference, because any differences in the eveness of the light will be theoretical and won't be visible.

The point about the recessed flash tube on the Godox AD600 is a good one. If you want to buy a flash that does its job well, get it from a manufacturer who understands about lighting, not from Godox:) Does anyone remember that awful flash head that Yongnuo introduced? The flash tube on that one was so deeply recessed that it was unusable...

Thanks Garry :)

And I agree absolutely with the distinction you make between say portraiture and high quality still-life/product photography where putting the light behind the subject and deliberately creating specular highlights where the actual light source is reflected in the subject. What's good for one may be useless for the other.
 
>SNIP
The point about the recessed flash tube on the Godox AD600 is a good one. If you want to buy a flash that does its job well, get it from a manufacturer who understands about lighting, not from Godox:) Does anyone remember that awful flash head that Yongnuo introduced? The flash tube on that one was so deeply recessed that it was unusable...
Ok.......recessed.........well then
profoto-b1-hypersync.jpg
:ROFLMAO:
 
I don't think I've ever seen a strobe that was completely even, there's always some kind of pattern to it. It just surprises me that a ring tube would project such a strong horizontal "bar" pattern, almost like the diffusion was curved/folded towards the head (but it doesn't look like it was). I would hope that's not actually characteristic of the D-lite...

If you ignore the two bright patches left and right explained above, the pattern is pretty regular. The other thing may be simple gravity, where the top and bottom surfaces of the softbox have a very slight curve (and it is very slight) not present on the vertical sides. I should have used an octa, but I don't have one without a central push-up mechanism that gets in the way.

I'm just a bit surprised by the pattern I saw like Steven mentions, shouldn't a square softbox produce a cross pattern rather than a single band with a standard exposed tube head? Gary makes a good point that people are going to be using that kind of softbox but it just caught me by surprise as it wasn't what I'd expect. Here's my 90x90 for reference:

View attachment 103815

I've been spoiled lately by rear firing octas, I'm guessing they'd eat too much light to be useful for a speedlight though (and far too big for convenience).

Having done this kind of thing before, I'm not at all surprised by the results, but maybe you have been spoiled by reverse-firing softboxes ;) On the other hand though, they kind of make the point why these apparent irregularities don't actually matter most of the time, ie reverse-firing softboxes always have a black circle in the middle which is the shadow of the flash unit in front, and with very big 6ft octas it's common practise for the photographer to literally stand in front of the light, blocking out a large part of it, but with minimal effect on the image.
 
Ok.......recessed.........well then
View attachment 103823
:ROFLMAO:

LOL Profoto often comes up in these debates, but the point is not how recessed or otherwise the flash tube is, but whether the light is able to fill the softbox freely and efficiently without obstructions. The main obstacles are usually the speedring, speedring adapter, and the general design of the mount, rather than the light unit itself. And of course Profoto has their 'zooming' mount so you can always push the light forward and clear of obstructions.
 
The internal of my softbox
View attachment 103818
Note no inner diffusor
Flash set at 24mmView attachment 103819

Wide angle reflector down (14mm)View attachment 103820

Wide angle diffusor and "stofen" diffusor
View attachment 103821

@24mm and stofen attached
View attachment 103822

Interesting, but what is the effect on the subject, and how do exposure levels vary?

Also, the wide-angle zoom head setting makes a difference, and the Yongnuo's 20mm performs better than your 24mm. At 20mm the flash is close to filling the front of a 90 degrees softbox, at least on the horizontal axis. That's easily fixed by using the wide-panel, but that costs precious brightness.
 
I should note that most softboxes have some degree of hot spot. And that's not necessarily a bad thing, in fact it can be very useful... I suppose that's why the inner diffuser is usually removable.
 
The point about the recessed flash tube on the Godox AD600 is a good one. If you want to buy a flash that does its job well, get it from a manufacturer who understands about lighting, not from Godox:) Does anyone remember that awful flash head that Yongnuo introduced? The flash tube on that one was so deeply recessed that it was unusable...

Would that be one of the Godox lights that lencarta sells?

Mike
 
Me too - and that is a key point!
I think it's more the fact that the pattern is pretty consistent... but there are still subtle differences. I.e. between 2-3 the catchlights are hotter, and the gradients (highlights and shadows) notably harder/denser in #3. It's kind of hard to notice looking at them individually though.
But if you were to layer them in PS and turn one on/off it would be obvious.
 
Interesting, but what is the effect on the subject, and how do exposure levels vary?

Also, the wide-angle zoom head setting makes a difference, and the Yongnuo's 20mm performs better than your 24mm. At 20mm the flash is close to filling the front of a 90 degrees softbox, at least on the horizontal axis. That's easily fixed by using the wide-panel, but that costs precious brightness.
Ill find something to shoot later but I doubt the differences will be dramatic.
My seconic 308B @ 1/125 (flash @ 1/8 power) centered 1m (GMP calibrated measuring tape :D ))away reads the forlowing.
24mm. F/4 + 4/10
24mm + stofen. F/4
WA diffusor + stofen. F/2,8 + 9/10
WA diffusor. F/4 + 2/10
 
Last edited by a moderator:
I think it's more the fact that the pattern is pretty consistent... but there are still subtle differences. I.e. between 2-3 the catchlights are hotter, and the gradients (highlights and shadows) notably harder/denser in #3. It's kind of hard to notice looking at them individually though.
But if you were to layer them in PS and turn one on/off it would be obvious.

For me though, and I suspect the vast majority, the question is does it matter...? No. But any gains in brightness with low-powered speedlights, that's worth chasing.
 
Good point... and that begs the question; are these images all flash (i.e. black exposure)? That would (should) make more of a difference.

Steven, please.
 
For me though, and I suspect the vast majority, the question is does it matter...? No. But any gains in brightness with low-powered speedlights, that's worth chasing.
A lot of the finer points don't often matter... If she was a real person w/ real hair it would be less apparent. If there was any other light it would be less apparent. If there was more BG separation it would be less apparent. And if it was just fill, it wouldn't really exist at all.
Even as it is, the differences are small enough to go unnoticed or be a matter of taste... and I'm guessing that if the speedlight was used from just a bit closer at a little lower output they would be practically identical.

But, I'm still torn by the very uneven fill w/ the D-lite... that's what's actually the equalizer in this test. I don't know what Simon is using in his 90, or which one it is, but if we were to compare that output to a speedlight in it I expect the results would be more notably different (but I've been wrong before).
 
Last edited:
Back
Top