How to do long exposures without ND filters. Techniques and examples

Long exposure is one of the most popular, yet complex,  techniques in landscape photography and the ND (Neutral Density) Filter is fundamental in reducing the light that hits the sensor thereby allowing the photographer to lengthen the exposure times. 

What many wonder is if there is a way to simulate the effect of the ND and therefore obtain the effect of a long exposures without using a filter? The answer is yes. 

The potential to simulate the effect of a long exposure shot is, initially , intriguing to many. Maybe because it is inexpensive or maybe for the technical challenge. However, in the long run, it is not sustainable. The effort required is disproportionate to the end result and most of all, digitally simulating a long exposure deprives you of the “soul” of the shot.

That moment when you have scouted the location, composed your shot, calculated your exposure times and then hit the shutter release and handed the photo over to forces beyond your control. 

What software to use to simulate long exposure

The software looked at here is Adobe Photoshop, we have not used anything else. There are no miraculous settings that can completely replace filters. When it comes to physics and optics there is no truly effective shortcut. 

Long exposure guide without filters

Step 1 is pretty straightforward; take the photo you have in mind. Once you are sure of the composition and exposure, make sure you manually set the focus, exposure and aperture so there are no variations in subsequent shots. Now you have to take many photos one after the other.

Many photos! 

The more photos you take, the closer the effect will be to a genuine long exposure shot.

The number of shots required is substantial though. To simulate 30 seconds of exposure, in theory you have to take 500-800 photos to replicate the effect (with a shutter speed of about 1/30). Obviously this is practically impossible, due to memory card capacity, the lag between one photo and another and the virtual impossibility of processing so many photos at the same time. So in the end you will have to settle for 30, 40, 50 or maybe 100 shots.

Once the shots are saved on your hard disk (we recommend an SSD to avoid biblical processing times) open Photoshop and select: File -> Script -> Upload file in series. Once the files are loaded, select Layer -> Advanced objects -> Image series method -> Arithmetic mean. And your photo is ready.

Using this method, which is the most advanced, you can also try taking fewer photos, about 40-50 about 1″ from each other. Or less photos about 3-4″ apart. The more photos you take, the better the result, theoretically you have to take infinite photos to have the same effect as an ND filter, which of course is impossible.

Unfortunately there is no right number of shots or the perfect interval, as usual everything “depends”. Now we are going to compare the long simulated exposure with the real one and see the possible defects.

Example of long exposures without ND filter

Examples are always the best way to show a concept, especially when it comes to photography. Below we have two shots, one simulated without any filters and combined in photoshop (left) and one with filters (right). There are about 50 photos taken in a burst and joined according to the arithmetic average method. In both shots, only the lens correction profile has been applied. No other alteration or post production has been made.

The images speak for themselves. Let’s start with the indirect advantages, which depend on the use of filters:

• The GND (Graduated Neutral Density) filter: The sky is really well defined, foreground is brighter and in general the photo on the right is more interesting, an excellent starting point for post production.
• The polariser has made its presence known, the grass has more vivid colours, and the water in the foreground is more transparent and with brighter colours.
• Finally we come to the Standard ND filter, the clouds are much more defined and with a very clear long exposure effect. But the effect on the water is the most obvious. With the simulation the water is not silky, but simply an overlap of many photos. The effect is not pleasing. 

We also note how NiSi filters have a very high definition, the photo does not lose sharpness, indeed, using image stacking in post production is virtually guaranteed to lose definition by comparison to using high quality filters such as NiSi. We also see the effect of the polariser on water, reflections and in general on the saturation of the image.

In this detail we see both the effect on the water and the transparency thanks to the polariser. In the simulated shot (left), the water is not very smooth, even if it is less noticeable near the shore. What is evident is the effect on the foam around the stone (practically absent on the simulation) and obviously the absence of the polariser effect.

In this detail of the shots we want to show another very important effect: the effect on water and trails. By simulating the ND filter (left) the photo has an effect ‘similar” to the silky effect we want, but it is not really smooth and also the trademark stripes do not emerge as when shooting the sea with a true long exposure. The whole atmosphere of the long exposure is lost and the result is very dull. Additionally, wave ripples remain visible when we attempt to simulate the ND filter.

Defects of the simulated long exposure

Obviously, like all “simulations”, it has its flaws. From the previous photos we have seen the major problems in the example shot. Now we will try to summarise the major problems both on shooting and in its management:

• You have to take many photos to get one simulated long exposure shot. And regularly shooting this many photos quick succession can have a detrimental affect on the life-span of the shutter (with an SLR). 
• Lots of shots require lots of space and processing capacity (HDD / SSD, processor, RAM, etc.). A RAW photo takes about 40 to 80MB, so if you want to try to simulate the long exposure you will have at least 3GB of shots for a single photo! And this only in the case of 50 shots of 60MB. If your RAWs are heavier and/or you take more photos,  you can easily find yourself with 6GB or more for a single photo. In addition, backup becomes a problem, especially when travelling. The processing power to upload all photos to Photoshop is also critical.  If in the end you are unable to process the photos it’s a pretty big problem. In processing the shot above we overloaded a top of the range MacBook Pro 15″ – it took us about 20 minutes to do all the processing and it was only 50 photos.
• There is no immediate feedback. Only once you have finished the uploading and processing will you know if the shot was successful? Did you take enough photos? Was the shutter speed correct? Ultimately you will fire off hundreds of shots just to be sure. This increases the problems of processing, storage and shutter wear. Equally, one of the benefits of digital photography is an instant response. If the shot is wrong, in any way, you can more often than not try again. But if you have fired off 50 shots, packed up and travelled home only to discover something is wrong, you’ve blown it. Effectively negating the whole benefit of digital photography. 
• Ghosting, which you won’t notice until it is too late. Imagine the passage of a boat on the sea, of a person, of a seagull or something else. When taking a true long exposure shot, these intrusions into the frame would be insignificant. They will result as a strip that can be left, contributing to the shot, or removed. And in most cases, especially with ultra-long exposures you won’t even see them because they have not remained in one point long enough for the sensor to pick them up. 
  But what happens when simulating the long exposure? Here the problem is huge! That seagull is now imprinted, in a different position, in every shot. So 50 shots, 50 seagulls. Now you HAVE to remove them, in every. single. frame. 
• The GND and the Polariser cannot be simulated, if you want to balance the shot (sky/background and foreground) and you want to do it with HDR or other techniques, this cannot be done (or is so complex it’s not worth the effort) when taking multiple shots for your “long exposure”.  In fact, HDR makes use of bracketing, that is, taking several photos with different exposures and then merging them. But this is not possible if we are taking a series of photos to simulate the ND. So we might as well add an ND to our slot if we also have a GND and polariser mounted. Using holders and filters (ND, GND and polariser) is the most complete, practical and best way to manage the light in the field and therefore obtain the best result in the best possible way in landscape or architectural photography.
• You lose all the beauty of photography in this field and maybe this is the biggest flaw of all. Landscape photography is beautiful because it allows you to experience shooting in the field, immersed in the landscape. The best results captured in-camera are always the most satisfying. 

Summary. Why the result is better with filters.

Ultimately, ND filters offer a better result than the simulated long exposure. But why? Think of a point (pixel) of the photo, on the sea for example. If you use an ND filter and expose for 1 minute this point will move with the typical movement of the sea and leave a trail that will be impressed upon the sensor for an entire minute. On the other hand, if we simulate the ND filter we will be shooting 10, 20, 50 or even 100 shots, but it will never be the same as continuous exposure for 1 minute. The point or pixel taken into consideration will be in 10, 20, 50 or 100 different positions depending on how many photos we have taken. The problem is that 50 points don’t make a line, only infinite points make it.

The example above tries to simply explain the difference between ND filter and software simulation. On the left we have several points, one for each photo that we took and merged together trying to simulate the ND filter, while on the right the moving point impressed a continuous trail upon the sensor thanks to the ND filter allowing us to keep the shutter open continuously. Obviously the concept has been simplified, but the result that you see in the photo comes from this. The ND filter is essential, unless you want to take tens of thousands of photos for an exposure of a couple of minutes (and it would still be an approximation!).

Is it worth simulating the long exposure?

If, like most photographers, you already have Photoshop it’s probably worth a try, just to satisfy your own curiosity.  Try to simulate a long exposure, it is certainly a very interesting method to become familiar with powerful software such as Photoshop. Above all, it is important to test firsthand the difference between a real photo, made in the field with quality ND filters, compared to an imposter that tries to simulate the effect. The results can be satisfactory in some cases, but if  you get home, load all the photos into Photoshop and the shot isn’t right, the magical moment for that shot has long passed. 

We recommend you read our guide to long exposure, with all the steps to get the best results and take full advantage of the fantastic passion that is photography.

Find out what you can do using NiSi filters: here you will find excellent examples of using filters, while we have several examples of using ND filters in landscape photography at this link. Finally, in the next link you will find excellent advice on which ND filter to choose.

Do you want more information? Any questions? Write to us!