tom_mann1

<p>... and it's histogram</p> <div></div>

<p>What makes you think you need to "fix it"? Sure, the image that produced your histogram has a preponderance of very dark and very light tones with few in the middle, but if the image itself looks good and serves your purpose, that's all that matters. </p> <p>For example, here's an image of some slightly out of focus black and white text and the associated histogram. Other than wishing it was a bit more in focus, I don't see any need to make tonal corrections.</p> <p>Tom M</p><div></div>

<p>I realize I'm stepping into the lion's den by making a comment about the 6.5 vs 8 fps issue, but I'll bet the guys who took iconic ringside boxing photos for Look and Life magazines using press cameras would have gotten a quite chuckle over this discussion (ie, vs their single shot, peak-of-action approach). </p>

<p>....argh. Rushing and forgot to attach the image.</p><div></div>

<p>If you haven't done so already, you might want to look into Topaz Adjust. With it, you can achieve results like this.</p> <p>Tom M</p>

<p>Why don't you post an example or two (before-after would be best) of the type of effect you are seeking to achieve.</p> <p>Tom M</p>

<p>PS - The purpose of the high pass filter on your mike is to minimize very low frequency mike handling noise, vibrations from the floor coming up through the mike, the very low frequency sound almost all air handling systems generate, etc. You won't hear these unless you are playing the recording back through a good sound system. </p> <p>The cutoff frequency of the HPF on your mike is well below the lowest frequencies put out by a singer and his guitar, so don't worry about it interfering with your recording -- just leave it on (should you wind up having to use your current mike).</p>

<p>IMHO, the way to essentially guarantee a good recording (both audio and video) is to separate the recording of the audio from shooting the video. This is probably a bit more complicated process than you expected but you can get superb results this way.</p> <p>First, get a really good quality audio recording without any thought to the video. Since you state that you don't know much about audio recording, find someone locally who does (...say, a member of a band), and has some basic equipment, ie, decent mikes, a simple mixer (and/or software), and access to an acoustically nice room (ie, reasonably large and broken up with as few parallel surfaces as possible) .</p> <p>One fairly standard way to mike a guy and an acoustic guitar in a good room is a pair of figure-8 studio mikes, about a foot or two in front of the guy and his guitar. The front lobe of one points at his mouth, the front lobe of the other towards his guitar. The rear lobes of both mikes pick up room ambiance / reverberation. Lots of nice tutorials about this setup, e.g., <p>If you aren't in an acoustically nice room, mike him like one would for a stage performance: one cardioid pointed to, and near his mouth, a second pointed towards his guitar. The exact placement of the guitar mike is always the subject of personal taste, style of music, etc.<br> <br />Once you have mixed it down and have a good audio recording, put him in an appropriately lit, visually nice environment without any mikes, play the recording back to him and have him play and sing along (ie, lip sync) while you are shooting video.</p> <p>Add the good audio to the video tracks, and presto, you have the best of both worlds.<br> <br />Just my $0.02,<br> <br />Tom M</p>

<div></div>

<p>Hey, Glen, thanks for posting those photos. They were very well done and interesting. To be honest, knowing the work I would have to go through to dig out my own notes and photos from our levitated droplet experiments that many years ago, I didn't even try, LOL. <br /><br />Given the subject matter, it sounds like there was a reasonable chance we might have overlapped at some conferences back in the day, although the work of my group was more directed towards understanding CBW aerosol clouds, not reactor / nuclear waste applications, and probably a decade earlier than yours.<br /><br />Anyway, take a look at some other photos of water emerging from shower heads. Note the presence of reasonably regular strings of droplets in all of these photos:<br /><br />http://janiceperson.com/wp-content/uploads/2013/03/9IMG_8755BW.jpg<br />http://img.alibaba.com/img/pb/725/893/239/1270026448706_hz_myalibaba_web_temp2_2794.jpg<br />http://us.123rf.com/450wm/janka3147/janka31471204/janka3147120400020/12997085-showerhead-and-falling-water-drops.jpg<br />http://www.patdolanplumbing.com/wp-content/uploads/shower-head.jpg<br />http://www.4freephotos.com/images/u/Shower-head-horizontal346.jpg<br />http://www.4freephotos.com/images/u/Round-shower-head-with-water1712.jpg<br />http://www.waterwise.org.uk/data/Waterwise_Images/Water_images/Adarshr.jpg<br />http://thumbs.dreamstime.com/z/close-up-water-flowing-out-chrome-shower-head-30844300.jpg<br /><br />In the late 1800's, Lord Rayleigh proposed a mechanism for these. The mechanism has been validated under a wide range of conditions and it now carries his name, the "Rayleigh instability" (http://en.wikipedia.org/wiki/Plateau%E2%80%93Rayleigh_instability). A short explanation for it is given in the 2nd section of the Wikipedia article. That discussion focuses on estimating the size of the resulting droplets, but once you have the size, indirectly, you also have an estimate of their spacing. <br /><br />Here are some more images that illustrate this phenomena. <br /><br />http://www-rohan.sdsu.edu/~rcarrete/teaching/M-596_patt/images/waterjet.jpg<br />http://www.aa.washington.edu/research/combustion/images/image008.jpg<br />http://www-rohan.sdsu.edu/~rcarrete/teaching/M-596_patt/images/waterjet_color.jpg<br /><br />So, yes, Barry, I'm quite sure these really are formations of droplets, not glints from oscillations of individual droplets, although the latter certainly will modulate (to varying degrees) the intensity of the brightness of each droplet in the photo.<br /><br />That being said, I can not explain all of the details in the image under question, particularly questions like why some of the tracks are long "dashes" whereas others are short drops, why some are pointed in different directions (...are they really droplets that have bounced), etc.<br /><br />This was definitely a good fun thread and great conversation. Let's hope that with the removal of the old off-topic section, the powers-to-be feel this thread has sufficient photography content so that they don't pull it for being more related to physics and engineering. <br /><br />Cheers,<br /><br />Tom M</p>

<p>I think you are pointing out that there are moderate deviations from regularity / equal spacing in any one trajectory. This, of course is true.</p> <p>However, the point is that the randomness in refraction / glints / diffractive scattering produced by shape oscillations of droplets is usually so extreme, you would never see anything looking even vaguely regular in the "dots and dashes" patterns that we are seeing in this photo.</p> <p>FWIW, back in the late 1980's I did some experimental work in shape oscillations of droplets acoustically levitated in a flow of surrounding air. We looked at droplets from about 100 microns up to a few mm. The goal was to see if such oscillations significantly enhanced evaporation of the droplets when compared to mass transport theories of the time which assumed no such oscillations. Enhanced mass transport would be important for modeling for both environmental issues and clouds of chemical agents.</p> <p>Cheers,</p> <p>Tom M</p>

<p>I can guarantee you that the equally-spaced-trail-of-dots effect seen throughout the cited photo is not due to shape oscillations of the droplets. </p> <p>The reason is that other than for the smallest droplets, all larger droplets support multiple modes of deformation (eg, oblate-to-prolate ellipsoid, "three leaf clover" modes, 4 lobed modes, etc. along one axis, as well as numerous circumferential modes around that same axis. Each specific combination of an axial and a circumferential mode will be at a different oscillation frequency. In the real world, none of these modes of oscillation are degenerate (ie, have the same frequency). Even worse, if the amplitude of the shape oscillation is large, the oscillations will be nonlinear (vs quasi-linear), and the resulting motion is almost always quasi-chaotic instead of being able to be represented by linear combinations of fundamental modes (aka, "eigenmodes").</p> <p>When a droplet is formed, or if it splits into two, or hits something, etc., etc. a huge variety of oscillation modes is excited 99.9% of the time. It fact, experimentalists in this field find it extremely difficult to prepare droplets in just one pure mode of oscillation (ie, one frequency) without using special techniques. The consequence of this is that the glinting of light from such droplets will almost always appear random, NOT equally spaced such as seen in many of the trajectories shown in the cited example.</p> <p>Tom M</p>

<p>One minor notational clarification: The unit of flash energy is watt-seconds (aka, Joules), not watt/second (ie, watts per second). The first is the integral of the instantaneous power (vs time) over the entire duration of the flash, ie, the total energy contained in the flash. </p> <p>The second, watts/second (note the replacement of the hyphen with the division symbol), is the rate of change of power over time. This unit is almost never used outside of engineering and physics circles, and even then, much less frequently. If you see it used in a photography lighting context, it is almost certainly an error.</p> <p>Tom M</p>

<div></div>

<p>MH: <em>"Maybe the evaporated-on metallic coating is 'lifting' by a few microns? As both are, in-effect, transparent, and they are semi-touching, I guess you could make Newton's Rings...."</em><br> <em> </em><br> I think we can eliminate that possibility. Evaporated metal coatings as thin as the ones used in such filters simply don't have the structural integrity to lift off and stay in one piece. Think of a piece of metal whose thickness is a tiny fraction of the thickness of gold leaf.</p> <p>However, it's possible that the applied metalization was not of uniform thickness and we are seeing a bit of interference between reflections from the front and back of the metal coating. The extremely high loss tangent (epsilon double prime) of the metal would make any such interferences very, very broad (in wavelength), but I suppose it could happen. Problems with non-uniform coatings are not all that uncommon, so one could have slipped through their QC checks.</p> <p>Tom M</p>

Your main camera fails when for the first time in 20 years, you forget to bring your backup to the wedding.

<p>Another thought ... in a lot of summer outdoor Bluegrass concerts, the stage is partially surrounded by a nearly white tent. This can be an absolute blessing because it provides wonderfully even, soft ambient light. It's almost like you are photographing the performers inside a light tent used in product photography. However, if the tent is colored, it's likely you'll get a color cast on the skin that won't be flattering and may be very difficult to correct to a satisfactory degree. In this case, let your flash provide most of the exposure -- don't just use it as a fill.</p> <p>Below is a photo taken around sundown at a local venue that is often used for Bluegrass performances. The tent over the stage is nearly pure white, and the lighting was entirely ambient. Note the nice rendering of the skin of the kids.<br> <br /><br /><br> Tom M</p><div></div>

<p>...closeups of instruments ...</p><div></div>

<p>...and audience shots ...</p><div></div>

<p>Lots of good suggestions. Here's a few more:</p> <p>1. If possible, get pix of the band during the sound check, or even an off-stage warmup / jam session. It will be earlier in the day (ie, lighter), and you will be able a lot closer, possibly, even on stage with them. This will give you a very different, more intimate perspective.</p> <p>2. Don't forget to turn around and get shots of the audience. There can be some great subjects at BG concerts, especially when the little kids get up and dance.</p> <p>3. The suggestion to take shots of individuals and pairs of musicians (especially, if they are interacting with each other) is excellent. Practically every BG song has an instrumental break, and often the players will look at, or nod to each other just before the break to silently indicate who is going to take this particular break. Learn the song structure so you can anticipate when the break is likely to occur, as it's a good time to capture the interaction between the band members. Also, take the suggestion to zoom in to the extreme and get tightly cropped shots of instruments being played.</p> <p>4. Unfortunately, the typically sized flash softener for hot-shoe flashes is much too small to have any effect when you are shooting outdoors at more than yard or two away from the subject. About the only effect it will have is to spread the light over a larger angle and drain your batteries faster, :-( .</p> <p>Tom M</p> <div></div>

<div></div>

<p>Let me clarify one aspect of this sort of posterization: The overall shape of the banding (ie, the nearly concentric rings) is caused by the use of 8 bits per channel, NOT the JPG compression, <em>per se</em>. The largest effect of JPG compression is that it causes roughness of the edges of the bands. In all cases, the banding effect is exaggerated by anything that increases either the local or global contrast in the effected part of the image.</p> <p>To illustrate this, I put together the following comparison. The left hand column shows the images that I started with. To make any banding that might be present easy to see, the right hand image in the same row is after I applied a strong "levels" adjustment to each of the LH images. This consisted of nothing more than moving the L and R sliders directly under the histogram inwards to increase the contrast. I used the same levels adjustment on all three rows of the table.</p> <p>The gradients were drawn using the gradient tool set to "radial", not by using a mask.</p> <p>The top row shows that if everything is done at 16 bits per channel (bpc), banding is virtually unnoticeable, even after the levels expansion. Note: if you have an ordinary 8 bpc (aka, "24 bit") graphics card or monitor, you may still see some banding, whereas if you have a 10bpc (aka, "30 bit") graphics card and monitor, and it is set up correctly, you should see almost no trace of banding in this row.<br /> <br />The middle row shows what happens if you do everything at 8 bpc: Banding is starting to become visible before levels expansion, and is very evident after levels expansion.</p> <p>The bottom row shows what happens if, instead of starting with a perfect gradient drawn with the gradient tool, one starts with a quality = 5 JPG JPG version of the same gradient. There is some minor increase in the visibility of the banding, but, by far, the largest effect of the JPG compression is roughening of the edges of each transition. This is because the JPG compression algorithm only operates on small (8x8) squares of pixels.</p> <p>HTH,</p> <p>Tom M</p> <p>PS - I posted the composite with a width and height under 700 pixels to try to avoid the tender mercies of the forum's image uploading software. If it was turned loose on these images, the effects illustrated would be muddied by the problems it introduces.</p><div></div>

<p>If anyone thinks there is any significant difference between an original 8 bpc RGB image and a version of that image that has been turned into a quality = 12 JPG in PS, the answer is simple: Please just perform the experiment for yourself with one of your own images. </p> <p>Put each version of your image on a separate layer in PS, and set the layer blending mode of the upper layer to either "difference" or "subtract". The result will perfectly black indicating no difference between the two. </p> <p>If you want to look for even miniscule differences between the two versions, put a "levels" adjustment layer above the pair, and slide the RH slider (directly under the histogram) almost all the way to the left, thereby magnifying any tiny differences that might be present. To convince yourself that this comparison technique actually does work, repeat the comparison with a quality level 7 (or below) JPG, and you will see lots of differences.</p> <p>There have been several threads on this comparison technique in photo.net in past years showing the above results, but I don't have the time at the moment to find them and provide a link.</p> <p>IMHO, the most important difference between the two file formats is the effect of bit rot over time. A single bit flipped somewhere in a JPG file almost always mucks up the remainder of the image after that point, and may even make the file difficult, if not impossible to open by conventional SW. </p> <p>In contrast, a single bit flip in an uncompressed TIF only impacts one of the three color channels in one pixel, nothing else. Such a small flaw in the image can either be ignored, or, if desired, can be easily repaired based on the content of the surrounding pixels.</p> <p>Put differently, the probability for a JPG to get significantly mucked up is hundreds of thousands, if not millions of times greater than that of an uncompressed TIF. It is the ratio of one bit, to roughly the overall size of the file in bits. This loss of robustness is so severe, that as I recall, the US Archives put out a position paper strongly suggesting archiving important images in uncompressed TIF or an archivally equivalent file format.</p> <p>HTH,</p> <p>TomM</p>

Sorry - I clicked on the ladder truck file instead of the one for the ambulance. Now, when I attempted to delete that post, I see that the edit/delete option to remove my earlier post is gone, so, mods, feel free to do so. Anyway,here's the correct file for a NW thread on emergency medical vehicles.<div></div>

<div></div>