Peaking Levels Accuracy at DVinfo.net

Chris Harding · December 28th, 2014, 08:11 AM

Hi Guys

I haven't done a test on this but does anyone know if a low level peaking setting would be more accurate in setting focus manually than a high level setting. I'm assuming here that the high level tends to enable more peaking compared to the low level? Or have I got it wrong??

Has anyone noticed a more accurate focus if you use a low level rather than a high level? I was thinking especially in situations where the iris is wide open so your DOF is more critical ... obviously at F16 in bright light your DOF is massive so almost everything is sharp but in very low light it's a lot more critical

Chris

Dave Baker · December 28th, 2014, 09:46 AM

I can't answer your question Chris, but if I may I would like to add a further question to your query.

Does anyone know whether picture profiles have any effect on peaking accuracy? Specifically, does using a flat profile make peaking less accurate than a higher contrast one?

Dave

Chris Harding · December 28th, 2014, 06:13 PM

Hey David

I think that part of the answer would be simple IF we knew whether peaking display is generated before the incoming signal is processed or not. However even if the peaks are generated pre-profile they surely will be less visible on an LCD that is showing a flat image as opposed to a high contrast one so even if the peaks are there a low constant display simply might not allow them to be visible.

My focus is razor sharp when using lighting at receptions, however now and again at ceremonies I do end up with a softer image and it's probably the combination of low light in a Church, plus no real hard edges with the couple ... peaking seems to be more accurate if you have hard corners like a lectern so my readings are always good, even in low light. Peaking I think finds soft folds of fabric tough to work with and then add a flat low contrast image and it might struggle even more?

Are you talking about peaking with Magic Lantern Dave??

Chris

Dave Baker · December 29th, 2014, 02:37 AM

Partly Chris.

I'm not trying to hijack your thread, I threw this in because peaking seems to be a bit of a black art and as yet I haven't found any satisfactory answers. As you know, it's easier to focus a high contrast image with lots of sharp edges by eye than it is a low contrast image with none. I believe it's the same for peaking and wondered if anyone here knows for sure.

I don't want to elaborate, that's new thread stuff.

Dave

Chris Harding · December 29th, 2014, 08:02 AM

Hi Dave

You are not hijacking at all ..it's all relevant!! Yeah in low light I think the peaking has the same issue as an auto focus system ..not enough contrast!! I was simply wondering if setting a lower peaking level might make focussing easier and was wondering what the level actually does...if you set it to high it tends to tell you that everything is in focus so the focus can easily be soft. I often find that in low light and low contrast the peaking tends to "tell" you to focus behind the subject but that's often an issue with DOF as there is always more depth of field behind the subject.

Funnily I seldom, if ever have any focus issues with the Sigma 18-35 as it's F1.8 so it's fast enough to resolve a better image in low light..... Hmmm maybe I need a focussed light on the camera that zaps the subject for a few seconds (prior to the ceremony of course) which will improve the contrast?? Hell, my still camera has a focus assist light so why couldn't one have a focus assist light on a video camera??

I wonder if it would work or if anyone has tried it?? Worth a test I think and see how peaking reacts to a low light subject and then what happens if we give the camera a brief high light subject.

Chris

Dave Baker · December 29th, 2014, 11:35 AM

Hi Chris,

Peaking Threshold or Percentage and presumably Level, sets how many pixels are considered to be in focus. The lower the setting the more accurate it is, but then it (the peaking line) can get so fine it all but disappears and it makes no difference what colour peaking you change to.

I did quite a lot of experimenting with this in Magic Lantern. I could set it up so that indoors, or out in the garden, I could instantly tell when I had focus, then walk down to the river or into the forest and, using the same lens, it all turned to sh*t. I used the split image in the focus zoom window where possible, but sometimes you just have to rely on peaking.

Dave

Steven Digges · January 1st, 2015, 02:14 PM

Hi Guys,

This is another "not for sure reply" but might help. Peaking has been available as a tool on monitors since the all CRT days and is still offered on many aftermarket LCD camera monitors. You activate it on the monitor so therefor it is reading the processed image.

I find the peeking on the EA50 to work better than any other peeking I have used. I love it. I cant answer your original post Chris but this is what I do.

I have settled on medium yellow. To me that is a LOT of peeking so I use the preset button to quickly turn it on or off. The EA50 peeking is so good at that setting you can see it throughout your DOF. I can see it more prevalent at the point of focus so I know where that critical point is. Then, when possible, I use focus magnification at 4X to confirm it. I find 8X useless as it magnifies grain, noise, pixels etc, too much.

For run and gun fast moving subjects I leave peeking on and have learned to trust it more than I ever expected to.

Low light, low contrast focus has always been difficult for us humans and the electronic tools available. Fortunately, if there is a face in the image, the most important point of focus often has some built in contrast if you are not too wide, EYES. The old school photographer in me is always concerned about eyes being in focus. If you never forget eyes you can find enough contrast in the eyes to focus. Even peeking shows up in the whites of the eyes pretty well. At least in some situations.

Steve

Chris Harding · January 1st, 2015, 05:39 PM

Thanks Steve!

Yep that's what I use as well .. yellow and medium. For weddings my A-Camera uses a very sharp Tamron 17-50 lens F2.8 and I use it for the ceremony and speeches mainly. Ceremony (being in a Church or outdoors) is done without extra lighting but speeches are done with a bit of bounced CFL light into an umbrella. The speeches footage is always razor sharp but ceremony now and again tends to look a wee bit softer
The eye tip also brings back my photography days so that's worth a try

Much appreciated

Chris

December 28th, 2014, 08:11 AM	#1
Chris Harding Inner Circle Join Date: Jul 2009 Location: Perth, Western Australia Posts: 8,441	Peaking Levels Accuracy Hi Guys I haven't done a test on this but does anyone know if a low level peaking setting would be more accurate in setting focus manually than a high level setting. I'm assuming here that the high level tends to enable more peaking compared to the low level? Or have I got it wrong?? Has anyone noticed a more accurate focus if you use a low level rather than a high level? I was thinking especially in situations where the iris is wide open so your DOF is more critical ... obviously at F16 in bright light your DOF is massive so almost everything is sharp but in very low light it's a lot more critical Chris

December 28th, 2014, 09:46 AM	#2
Dave Baker Major Player Join Date: May 2014 Location: Brandon, England Posts: 459	Re: Peaking Levels Accuracy I can't answer your question Chris, but if I may I would like to add a further question to your query. Does anyone know whether picture profiles have any effect on peaking accuracy? Specifically, does using a flat profile make peaking less accurate than a higher contrast one? Dave

December 28th, 2014, 06:13 PM	#3
Chris Harding Inner Circle Join Date: Jul 2009 Location: Perth, Western Australia Posts: 8,441	Re: Peaking Levels Accuracy Hey David I think that part of the answer would be simple IF we knew whether peaking display is generated before the incoming signal is processed or not. However even if the peaks are generated pre-profile they surely will be less visible on an LCD that is showing a flat image as opposed to a high contrast one so even if the peaks are there a low constant display simply might not allow them to be visible. My focus is razor sharp when using lighting at receptions, however now and again at ceremonies I do end up with a softer image and it's probably the combination of low light in a Church, plus no real hard edges with the couple ... peaking seems to be more accurate if you have hard corners like a lectern so my readings are always good, even in low light. Peaking I think finds soft folds of fabric tough to work with and then add a flat low contrast image and it might struggle even more? Are you talking about peaking with Magic Lantern Dave?? Chris

December 29th, 2014, 02:37 AM	#4
Dave Baker Major Player Join Date: May 2014 Location: Brandon, England Posts: 459	Re: Peaking Levels Accuracy Partly Chris. I'm not trying to hijack your thread, I threw this in because peaking seems to be a bit of a black art and as yet I haven't found any satisfactory answers. As you know, it's easier to focus a high contrast image with lots of sharp edges by eye than it is a low contrast image with none. I believe it's the same for peaking and wondered if anyone here knows for sure. I don't want to elaborate, that's new thread stuff. Dave

December 29th, 2014, 08:02 AM	#5
Chris Harding Inner Circle Join Date: Jul 2009 Location: Perth, Western Australia Posts: 8,441	Re: Peaking Levels Accuracy Hi Dave You are not hijacking at all ..it's all relevant!! Yeah in low light I think the peaking has the same issue as an auto focus system ..not enough contrast!! I was simply wondering if setting a lower peaking level might make focussing easier and was wondering what the level actually does...if you set it to high it tends to tell you that everything is in focus so the focus can easily be soft. I often find that in low light and low contrast the peaking tends to "tell" you to focus behind the subject but that's often an issue with DOF as there is always more depth of field behind the subject. Funnily I seldom, if ever have any focus issues with the Sigma 18-35 as it's F1.8 so it's fast enough to resolve a better image in low light..... Hmmm maybe I need a focussed light on the camera that zaps the subject for a few seconds (prior to the ceremony of course) which will improve the contrast?? Hell, my still camera has a focus assist light so why couldn't one have a focus assist light on a video camera?? I wonder if it would work or if anyone has tried it?? Worth a test I think and see how peaking reacts to a low light subject and then what happens if we give the camera a brief high light subject. Chris

December 29th, 2014, 11:35 AM	#6
Dave Baker Major Player Join Date: May 2014 Location: Brandon, England Posts: 459	Re: Peaking Levels Accuracy Hi Chris, Peaking Threshold or Percentage and presumably Level, sets how many pixels are considered to be in focus. The lower the setting the more accurate it is, but then it (the peaking line) can get so fine it all but disappears and it makes no difference what colour peaking you change to. I did quite a lot of experimenting with this in Magic Lantern. I could set it up so that indoors, or out in the garden, I could instantly tell when I had focus, then walk down to the river or into the forest and, using the same lens, it all turned to sh*t. I used the split image in the focus zoom window where possible, but sometimes you just have to rely on peaking. Dave

January 1st, 2015, 02:14 PM	#7
Steven Digges Trustee Join Date: Oct 2002 Location: Phoenix, AZ Posts: 1,828	Re: Peaking Levels Accuracy Hi Guys, This is another "not for sure reply" but might help. Peaking has been available as a tool on monitors since the all CRT days and is still offered on many aftermarket LCD camera monitors. You activate it on the monitor so therefor it is reading the processed image. I find the peeking on the EA50 to work better than any other peeking I have used. I love it. I cant answer your original post Chris but this is what I do. I have settled on medium yellow. To me that is a LOT of peeking so I use the preset button to quickly turn it on or off. The EA50 peeking is so good at that setting you can see it throughout your DOF. I can see it more prevalent at the point of focus so I know where that critical point is. Then, when possible, I use focus magnification at 4X to confirm it. I find 8X useless as it magnifies grain, noise, pixels etc, too much. For run and gun fast moving subjects I leave peeking on and have learned to trust it more than I ever expected to. Low light, low contrast focus has always been difficult for us humans and the electronic tools available. Fortunately, if there is a face in the image, the most important point of focus often has some built in contrast if you are not too wide, EYES. The old school photographer in me is always concerned about eyes being in focus. If you never forget eyes you can find enough contrast in the eyes to focus. Even peeking shows up in the whites of the eyes pretty well. At least in some situations. Steve __________________ www.CorporateShow.com Been at this so long I'm rounding my years of experience down...not up!

January 1st, 2015, 05:39 PM	#8
Chris Harding Inner Circle Join Date: Jul 2009 Location: Perth, Western Australia Posts: 8,441	Re: Peaking Levels Accuracy Thanks Steve! Yep that's what I use as well .. yellow and medium. For weddings my A-Camera uses a very sharp Tamron 17-50 lens F2.8 and I use it for the ceremony and speeches mainly. Ceremony (being in a Church or outdoors) is done without extra lighting but speeches are done with a bit of bounced CFL light into an umbrella. The speeches footage is always razor sharp but ceremony now and again tends to look a wee bit softer The eye tip also brings back my photography days so that's worth a try Much appreciated Chris