Shooting over ISO 2000: Choices of ISO, shutter speed, f stop and workflow.

[Asher Kelman]

In the original thread, here, orchestra pictures were part of a celebratory photoessay.

Here, however, we can address the technical aspects of shooting without flash of the players set on a stage lit unevenly from above. The pre-concert pass time is about a minute. Entire orchestra shots from below hide the woodwind, violas, percussion and more. One want's natural positions and obvious truth that they are truly playing music not pretending.

So here we can discuss choices and consequences of using higher ISO and other technical decisions and then workflow. We can look at how such live music performances can be better covered. This includes practical, budget-sensitive, additional portable continuous lighting to take on location to use from the front to tame ghoulish shadows on prominent facial features. (Flash would be possible at the end for the bows and ovations). Post-processing, color correction of uneven poor overhead lighting, shadows of spectacles and facial features and dealing with clutter have to be addressed.

Asher Kelman: Percussionist Sean Case warms up on the Glockenspiel
5DII ISO 4000, Tungsten above, 1/160, f2.0, 50 1.2L

Asher Kelman: Harpists Kayla Torrez Chang (left) and Charissa Barger (right).
Later they provided the spirited glissandi and mandolin-like trills during Ravel's Alborado Del Gracioso
5DII ISO 2500, 580 EXII, 1/160, f1.8, 50 1.2L

Asher Kelman: Untitled
5DII ISO 4000, Tungsten above, 1/160, f2.0, 50 1.2L

Asher Kelman: Flautist
5DII ISO 4000, Tungsten above, 1/160, f2.0, 50 1.2L
​

 

[Asher Kelman]

The Oboe!

Asher Kelman: The oboe in Art Jennifer Johnson's Hands
5DII ISO 4000 Tungsten above, 1/250, f4.0 70-200 2.8L at 160mm​

​

 

[Asher Kelman]

Preparation and stealth in shooting an entire orchestra...oh yes, and pixels!

Asher Kelman: Colburn Yehuda Gilad, Conducting, Conservatory Playing Symphonic Metamorphosis of Themes by Paul Hindermith
Canon 5DII, 50 1.2L at f4.0, 1/160 sec, ISO 4,000, overhead tungsten light, jpg, min processing in CS2​

Yes, I'll work on the faces in the shadows in the back. The prints will be made from the RAW file, of course.

 

[Asher Kelman]

Asher Kelman: Ronald Leonard, Cellist, left with Yehada Gilad conducter, Colburn Conservatory Orchestra, September 26th 2009, Santa Monica
Ron is accepting standing ovation for his performance of Concerto in B minor for Violincello and Orchestra, Antonín Leopold Dvořák, (1841-1904)

Canon 5DII, 70-200 2.8L IS at f4.0, 1/250 sec, 70mm, ISO 4,000, overhead tungsten light

Asher Kelman: "Colburn Cello Section adding their own ovation for Leonard's Dvořák!"
The Colburn Conservatory Orchestra, September 26th 2009, Santa Monica

Canon 5DII, 70-200 2.8L IS at f4.0, 1/250 sec, 105mm, ISO 4,000, overhead tungsten light, ​

 

[nicolas claris]

Hi my friend Asher
I'm back now and can see your pictures on another screen than the Iphone's!
Very nice work done in difficult conditions…
Some questions rising to me:
On my home laptop, all pics seems reddish, specially in the shadows… is it my monitor (this one not calibrated…)? or the ACR rendering which I always found having reddish shadows?
I also would be curious to see some crops at 100%, using ISO 4 000 is so amazing,I'd love to have that on my 1Ds3!
I noticed also the (relatively) high shutter speed, in such conditions I may have tried lower ISO and speed or lower speed and higher ƒstop (for larger dof)…

Will you have a chance to shoot within the stage during a training session?

Just some questions… so happy that you post your own work! a (too) rare privilege for all of us ;-)

 

[Asher Kelman]

Hi my friend Asher
I'm back now and can see your pictures on another screen than the Iphone's!
Very nice work done in difficult conditions…

Thanks so much! I value, as always, your interest and feedback!

Some questions rising to me:
On my home laptop, all pics seems reddish, specially in the shadows… is it my monitor (this one not calibrated…)? or the ACR rendering which I always found having reddish shadows?

I admit to adjusting the color based on white of shirts. I use the software plugin, iCorrect Editlab Pro, from Pictocolor. It's optimized, here, for correcting skin. First the white shirts are used as a reference to get rid of the contaminating hues. Subsequently in the final step, the skin is sampled and referenced to a library of skin tones. Then, as is my custom, I blend the masked color-corrected layer with the original. I may do several different color-corrected masked layers, for example, to correct the more yellow-hued hands. These receive stronger direct overhead light. The pictures are from jpgs as CS2 on my now old-fashioned Mac G5), can't handle the 5D II RAW files and I find the Canon RAW converters interface clumsy. JPGs do give immediate access but working on them instead of the RAW file is such a waste of time when they are going to be printed down the road! I need a new computer for my workflow that will take CS4; all offers kindly accepted!

Anyway, nicolas, thanks for drawing attention to the question of a red hue in the shadows. Bart may have ideas on this. We'll recalibrate my monitor, convert the RAW files on my laptop in CS4, and then look at the pictures afresh.

I also would be curious to see some crops at 100%, using ISO 4000 is so amazing,I'd love to have that on my 1Ds3!

I'll do that!

I noticed also the (relatively) high shutter speed, in such conditions I may have tried lower ISO and speed or lower speed and higher ƒstop (for larger dof)…

I found in the end, that with the 50 mm focal distance, 1/160 was a good speed for the whole orchestra at f4.0. The 50 mm 1,2L, is, of course a not motion corrected. At that aperture, with the great distance used from the balcony to the orchestra, the DOF is pretty good. With the shot of the orchestra playing, I needed to stop the fast movement of so many players. It's not often possible to shoot over again. So I have 1/160 as my lowest speed hand held at 50 mm.

I must admit, my shooting style for individuals is generally wide open as possible to blur cluttered background, especially on stage. I can see that in these shots I might be able to do down to ISO 2500, and sometimes I did that.

Will you have a chance to shoot within the stage during a training session?

That's the next plan! Then, we can deal with better lighting. We're looking into that. Traditionally, orchestras, unlike theaters, seem to be lit from above, shadowing the facial features!

Just some questions… so happy that you post your own work! a (too) rare privilege for all of us ;-)

Thanks again Nicolas. I'll post more.

Asher

 

[fahim mohammed]

Asher, I just got a chance to see these pics. I too have the following observations:

1. excellent high iso results on the monitor.

2. why the reddish hues? i have read Nicolas query and your response. was the lighting somehow affecting your wb?

3. I would have liked to see more close-ups like the beautiful OBOE shot. lovely.

4. you do like your 50/1.2 L..heavy beast that.

Although I can see the story telling in your series...overall, closer, very close etc. I wonder why you
were using such high iso/s? for speed, aperture, dof.

I would like to hear your impressions of how the prints came out.

Thanks for sharing what must have been a memorable performance.

Regards.

 

[Asher Kelman]

Another Stage lit unevenly from above with tungsten! 5DII, ISO 2000-2500, no flash!

This time it's the Chamber Music Recital at the Colburn School itself. The strings were showcased. It was a superb performance with sounds that surpassed anything I've heard at any major big city concert hall. This is likely due to the virtuosity, leadership, interpretation, technique, discipline and stamina taught under the influence of the Colburn String Faculty but in particular, Robert Lipsett and his own protegés who are now faculty teachers too. It was a divine experience that made thinking of photography difficult and I put the camera down for the duration. The music must have made the angels in heaven stop whatever they were doing. The man with the sickle was transfixed and I'm sure we'll all live 3 hours longer because of this.

I have tried another stage with a little more light using high ISOs of 2000 and 2500 and needed to correct the pictures for the shadows. Two needed 0.7 EV exposure compensation to prevent skin from just being blown out. But here are images I was able to deliver. These are quickly made from jpgs. The RAW would help of course but the exposure with better lighting is the key to future work.

f 4.0 1/166, -0.7 EV, ISO 2500

f 4.0 1/332, -0.7 EV, ISO 2500

f 4.0 1/256, -0.0 EV, ISO 2000​

 

[Bart_van_der_Wolf]

Hi Asher,

Your description of the performance makes me wanting to be there ...

I have tried another stage with a little more light using high ISOs of 2000 and 2500 and needed to correct the pictures for the shadows. Two needed 0.7 EV exposure compensation to prevent skin from just being blown out. But here are images I was able to deliver. These are quickly made from jpgs. The RAW would help of course but the exposure with better lighting is the key to future work.

The JPEGs look good enough as presented, however ..., just a tip for a Raw processing workflow, in case you haven't tried (and it can be tested at home). Set the camera at ISO 800 (maybe 1600, I don't know the 5D2 well enough) and underexpose the image by 1, 2, and 3 stops. It's easiest to use M exposure mode. The JPEGs will be useless, but the Raw data has probably lower noise than when shooting at an equivalent elevated ISO (=gain) setting. Of course the 'exposure' will have to be boosted at Raw conversion.

The potential benefits would be; no risk of blown highlights(!), and lower noise because the camera's amplifier noise contribution is reduced. Boosting the exposure in software doesn't create noise, while doing it in the camera with ISO does. You'll have to see how many stops you can underexpose and recover in your Rawconverter (Capture One allows 2.5 stops). You compare the ISO 800 and 1 stop underexposed shot with the ISO 1600 shot, 2 stop under exposed with ISO 3200, etc. You can try the same with ISO 1600 and underexposing, but I expect it won't be better than ISO 800 + an extra stop of Raw exposure boost.

I've read some reports about pattern noise on the 5D2 with low light exposures. If you have any hor/ver patterns showing in the shadows, then there are methods of reducing them, but it may get a bit too involved to do that routinely. You would also have to shoot a number of 'Dark frames' at the same exposure time as your other exposures.

Cheers,
Bart

 

[Asher Kelman]

Bart,

Does the shot for sensor noise have to be taken there and then, or at the same room temperature and settings? I've never used that and wonder what the function is. To do it in the middle of shooting an event? Does it simply add one dark frame for each shot?

Asher

 

[Bart_van_der_Wolf]

Bart,

Does the shot for sensor noise have to be taken there and then, or at the same room temperature and settings?

Same temperature, ISO, and exposure time, that's all that's needed for a Dark frame. The shots can be taken with the body cap on and viewfinder blocked. In principle the shots (say 8 of them) should be taken at approx. the same time as the actual exposures, to avoid sensor changes over time. However, we're not talking about rocket science, so it's okay to do them at a different time.

It is not certain that a given camera suffers from pattern noise, nor is it always disturbing the image. Only in large uniform and usually dark areas will it be potentially visible. But it won't hurt to determine its significance either.

I've never used that and wonder what the function is. To do it in the middle of shooting an event? Does it simply add one dark frame for each shot?

If you don't notice any regularity/pattern in the noise, then it may not be necessary to deal with it. If you do notice pattern noise, it's comforting to know that there is a remedy for Raw files, although it'll require special software (e.g. ImagesPlus or IRIS).

There are cameras (e.g. Sinar) that apparently take occasional dark frames over the course of a shoot, and use that at Raw conversion of files with the same ISO and exposure time. That would automatically take care of the 'same temperature' requirement.

Cheers,
Bart

 

[Asher Kelman]

Same temperature, ISO, and exposure time, that's all that's needed for a Dark frame. The shots can be taken with the body cap on and viewfinder blocked. In principle the shots (say 8 of them) should be taken at approx. the same time as the actual exposures, to avoid sensor changes over time. However, we're not talking about rocket science, so it's okay to do them at a different time.

It is not certain that a given camera suffers from pattern noise, nor is it always disturbing the image. Only in large uniform and usually dark areas will it be potentially visible. But it won't hurt to determine its significance either.



If you don't notice any regularity/pattern in the noise, then it may not be necessary to deal with it. If you do notice pattern noise, it's comforting to know that there is a remedy for Raw files, although it'll require special software (e.g. ImagesPlus or IRIS).

There are cameras (e.g. Sinar) that apparently take occasional dark frames over the course of a shoot, and use that at Raw conversion of files with the same ISO and exposure time. That would automatically take care of the 'same temperature' requirement.

Bart,

I'll have to read up on the use of dark frames with the 5DII. Is there a capability in any RAW processing software for using this for batch work?

In the meanwhile, I used your advice and underexposed at ISO 1600 and below to get rid of the burnt out arms and shoulders of violinists and cellists. It looks like this might be a good approach, at least from the image on the LCD. I'm thinking that perhaps I should use the Fuji DSLR which has a dynamic range of 14 ev as that should cover my needs.

The Nikon D700 also has a larger DR than the 5DII. Unfortunately, for orchestra pictures, neither of these cameras has enough pixels for the entire 116 people on stage. Still, your method of using ISO 1600, (or even less) with 1-21/2 stops of underexposure may have given me a great improvement.

Thanks for the generous guidance.

Asher

 

[Mike Shimwell]

Asher

Not having access to the giddy heights of iso 6400+ I have often used iso1600 or 3200 and underexposed by one stop as the resulting file is not (usually, provided the next step up is software based) significantly noisier and you gain extra highlight headroom. On occasion I have just used the slowest shutter and widest aperture I dared and relied on pst to bring an iso 3200 capture up by a couple of stops. It's noisy, but remarkably robust at a 10 by 8 print size.

Mike